|
Hi all,
I m working on Server - Client type application,using SOCKET to communicate between Server and Client.
I m using RegConnectRegistry to access client registry,but some times this function fails.
please help me for this.
i think its a client server connectivity issue because of security hierarchy.
it happens when server is 2003 and client is VISTA with uac enabled.
or Server is 2008 and client is XP.
please help me.
and please tell me how can use user name and password from server side to access the client.
provide me any help or guidance for this.
thanks in advance.
To accomplish great things, we must not only act, but also dream;
not only plan, but also believe.
|
|
|
|
|
When server is Windows server 2008 and clinet machine is Windows xp Professional with service pack 3.
in network window when i want to access this client machine its popup a username and password window to access.
in my code if i use RegConnectRegistry without entering username password in network window for xp client,this return a error of Access Denied.
if i use RegConnectRegistry after entering username password in network window for xp client,
than its execute,no error return but after the i want to open the registry by RegOpenKey.
now this return error access denied.
please help me what can i do.
thanks.To accomplish great things, we must not only act, but also dream;
not only plan, but also believe.
|
|
|
|
|
Hi All
I am try to get data from oracle.I am facing some problem to get data from oracle table when oracle table data type is RAW.For example i have a table in oracle
create table Test(Item RAW(255));
insert into Test values('1');
insert into Test values('2');
Now i am fetching data from Test table.
CString tabledata=Select * from "system"."Test";
pRecordset=m_pConn->Execute(_bstr_t(tabledata), 0, adCmdText);
_variant_t fetchdata;
fetchdata= pRecordset->GetRows(adGetRowsRest);
LONG lend = fetchdata.parray->rgsabound->cElements;
COleVariant varField;
for (int i = 0; i < lend; i++)
{
int k=0;
for (int j = 0; j < nCount; j++)
{
long index[2] = {j,i};
SafeArrayGetElement(fetchdata.parray,index,&varField);
}
}
out put is
varField = {???}
when i click on varFiled then show like this
tagVARIANT = safearray of UI1 = [1](1 '')
But original values is "1".
So how can i get original values.
Please help me
|
|
|
|
|
Hi,
We have SDI application. In which we are opening some dialogs. Once application is up these dialogs can be open and close as many times as possible. If I am opening dialog first time, minimize application(not dialog, i mean whole application) and then maximize application, dialog is also displayed in previous condition. But if i close and open dialog again, on minimize and maximize application make dialog disappear, I have to open dialog explicitly. How can I stop this behavior?
I tried to explain situation as much as possible. If anyone have doubt fire your question.
Thanks.
|
|
|
|
|
I don't know. Two questions do immediately come to my mind, so I will share them in case they lead you somewhere.
Are you consistently setting the parent window everywhere you are creating/recreating the dialogs?
(Presuming you are using MFC) How are you managing the lifetime of CDialog derived object associated with the dialog?
I hope these might lead you somewhere, but I don't know if they will.
|
|
|
|
|
First thing, I am not recreating dialog every time. I mean Once it is created i am not destroying its handle/object till application is down. So setting of parent is done once. I think this will be ans or your second que also. And ya I am using MFC.
|
|
|
|
|
Well, that sounds right. So you will be intercepting user request to close the dialog and hiding it instead.
Something has to have changed in the state of the dialog or object, but I'm afraid I'm not going to come up with it.
My remaining thought is to monitor the windows messages being sent to the dialog in the 2 situations. A difference would obviously be a clue. Failing that, it might suggest a good point to put in a breakpoint.
Good Luck.
|
|
|
|
|
No I am not stoping dialog from being close check this part;
void CHWSimulatorDlg::OnSysCommand(UINT nID, LPARAM lParam)
{
switch(nID)
{
case SC_MAXIMIZE:
ShowWindow(SW_SHOWMAXIMIZED);
bMaximizeFlag = true;
break;
case SC_RESTORE:
case SC_CLOSE:
bMaximizeFlag = false;
CDialog::OnSysCommand(nID, lParam);
break;
default:
CDialog::OnSysCommand(nID, lParam);
break;
}
}
I tried to debug this part, I am not getting any message in this function when I am minimizing whole application.
And I havnt override OnClose function for this dialog.
|
|
|
|
|
I don't think that the code you posted is the problem. It also doesn't seem that useful to me.
Now chevu wrote: No I am not stoping dialog from being close
Combined with other things that you have said, this suggests to me that we may actually be getting somewhere.
First, I'll explicitly state my understanding that we are discussing modeless dialog boxes with associated classes derived from CDialog.
Now let me try to quickly summarize lots of stuff. Some of this may seem to simple to mention. I'm doing it anyway because I suspect that you are missing part of this, but I don't know what part.
In working with modeless dialog boxes in MFC there are 2 things around that one has to worry about. One is a C++ object instantiated from a class derived from CDialog. This is normally dynamically allocated so that its lifetime can be properly managed. The other is the windows object (not a C++ object) that Windows maintains and deals with that a windows handle is connected to. They are associated, but they are somewhat independent things.
A class for a modeless dialog overrides OnOK and OnCancel and does NOT call the default versions. The default ones in CDialog call EndDialog() which is intended for use with modal, not modeless dialogs. (I'm not sure of all the implications of calling it for a modeless dialog, but I suspect this may be a part of your situation.) These overrides might call DestroyWindow() to terminate the dialog and kill the windows object. When it is desired that the dynamically allocated C++ object self-destructs when the dialog is closed, PostNcDestroy() is also overriden.
Typically, one constructs the C++ object and calls Create() which creates and associates the windows object. Later, the user exits the dialog in some way. The programmer has provided code for this that includes a call to DestroyWindow(). Then some programmer provided mechanism deletes the C++ object.
Now you want to do things differently. Once things are created, you want to keep things around (though not necessarily displayed) until the program exits. So when the user "closes" the dialog instead of destroying the window object (via DestroyWindow) and deleting the C++ object, you are writing code to hide the window and keep track that you have done this. Then when you want to display the dialog again you would just show the window.
It now sounds like this isn't what you are actually doing. Perhaps you are letting EndDialog() be called. I don't know just what state that would leave the window object in.
Hopefully this gives you some direction. Now I have to get to bed before I finish falling asleep.
|
|
|
|
|
Hi
I want to save class instance addresses to a vector. How can I get address of a class instance?
class A {}
A* pA = new A;
How can I get "pA" address value and assigned to a variable?
Thanks,
|
|
|
|
|
|
Thanks,
So I can implement like:
std::vector<void*> classAddress;
classAddress.pushback(&pA);
|
|
|
|
|
No, I don't think so - in that code pA is the instance pointer. &pA is the address of the pointer pA that holds the instance address, and is temporary (unless it's a global).
Instead, I would suggest:
class A { ... };
A* pA = new A;
std::vector<void*> classAddress;
classAddress.push_back(pA);
Why std::vector<void*> instead of std::vector<A*> though? Are you going to store pointers to different types there?
|
|
|
|
|
Thank you for your reply.
I need to save all kinds of classes. Not only class A.
|
|
|
|
|
In this approach, you will have problems knowing what each item is or being able to do anything with them.
If these objects are all related by all being the same sort of thing and based on a common base class, you could use a vector of pointers to the common base class.
If these objects are not related, you might think about another organization that will reflect what they are.
As a side note, don't forget about deallocating whatever you have allocated.
|
|
|
|
|
Avi Berger wrote: n this approach, you will have problems knowing what each item is or being able to do anything with them.
I don't do anything. I just need the address.
Avi Berger wrote: f these objects are all related by all being the same sort of thing and based on a common base class, you could use a vector of pointers to the common base class.
Because I need to do anything with the address, I think there is no need for a common base class.
Thanks
|
|
|
|
|
First, you need a semi-colon at the end of your class definition.
class A {};
Next,
transoft wrote: A* pA = new A;
pA is a variable that you have assigned the address of a dynamically created instance of A. So you have already accomplished what you were asking about.
So from your other message it would be:
std::vector<void*> classAddress;
classAddress.pushback(pA);
However, this will raise peoples hackles as you generally don't want to play with void *'s if yo don't have to.
Perhaps you might ask about what it is you hope to achieve.
|
|
|
|
|
Good afternoon
Please I want to subclassing dos commands but I don,t know how can I proceed.
|
|
|
|
|
aurelcly wrote: subclassing dos commands
"One man's wage rise is another man's price increase." - Harold Wilson
"Fireproof doesn't mean the fire will never come. It means when the fire comes that you will be able to withstand it." - Michael Simmons
"Man who follows car will be exhausted." - Confucius
|
|
|
|
|
huh?
This signature was proudly tested on animals.
|
|
|
|
|
aurelcly wrote: I want to subclassing dos commands
Not possible I'm afraid.
MVP 2010 - are they mad?
|
|
|
|
|
Try to explain what you mean, as you can see we don't quite get what you mean. Do you maybe mean you want to replace dos commands with your own versions, like, writing your own terminal or such?
> The problem with computers is that they do what you tell them to do and not what you want them to do. <
> Sometimes you just have to hate coding to do it well. <
|
|
|
|
|
Are you actually dealing with a genuine DOS environment or are you talking about a Windows console? I don't know the low level details of the Windows console, but I would not assume that they are the same.
Things that you might mean, but we can't tell which, if any:
1) Hook a DOS call. (presumably some int 21h function)
2) Replace an "external" DOS command. (provide a replacement executable earlier on the path)
3) Replace an "internal" DOS command. (I don't think this is going to happen without replacing the command interpreter, but perhaps I'm wrong.)
|
|
|
|
|
Right! INT 21H and such. Sounds quite distant.
I guess he is not interested in that type of things. He probably wants to mess with the Windows Console.
|
|
|
|
|
loyal ginger wrote: I guess he is not interested in that type of things. He probably wants to mess with the Windows Console.
I'd presume so as well.
It's not quite so distant for me though. I do some maintenance on an embedded system that uses a board support package that is a partial DOS work alike. I've had to reverse engineer part of its INT 21H implementation to find and patch some bugs in it.
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
